1. Technical Field
The present invention relates to an approach for searching both structured and unstructured data sources in a fashion that finds commonalities between the sources and exploits such commonalities recursively.
2. Description of the Related Art
Information users, such as governmental organizations, corporations, and other private organizations, typically manage both unstructured and structured data sources. Some of these sources may be maintained in-house, while others may be maintained, and therefore controlled, by external entities. The structured and unstructured sources often have little or no synergy, linkage, or automatic notification between each other. Mining data between these disparate types of sources traditionally falls short. One approach is to index the structured sources, however this approach ignores logic that can be applied based on structured metadata. A second approach is extracting structured data from unstructured data, however automated extraction technology has been researched for many years and yields structured data of questionable quality.